JPH07101394B2 - Automatic matching method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial field of application] The present invention is used for an information processing apparatus.

The present invention relates to an automatic collation system for large-capacity master files using a dedicated collation processing device.

〔Overview〕

According to the present invention, in the automatic collation system, the master data amount is distributed based on the characteristic information of the inquiry data to the master files of the same number as the number of the collation processing devices so as to be almost even, and the master files corresponding to the inquiry data are distributed. By adding contrast area information, the collation response time of the inquiry data can be shortened in both the online real-time method and the remote batch method.

[Conventional technology]

Conventionally, the automatic collation method for large-capacity master files using a dedicated collation processing device divides the master file into areas in a hierarchical structure, divides the divided areas into groups according to the number of collation processing devices, and All the collation processing devices were distributed to some of the collation processing devices, and a plurality of inquiry data were collated by all the collation processing devices at the same time.

[Problems to be solved by the invention]

However, in such a conventional automatic collation method, in order to speed up the collation response time, a large-capacity master file is divided into areas according to the characteristics of the inquiry data, and the divided individual areas are divided by the number of collation processing units. They were divided into groups and assigned to the magnetic disk devices connected to each collation processing device. Then, one inquiry data was distributed to some of the verification processing devices of all verification processing devices, and multiple verification data were verified by all verification processing devices at the same time. There was a drawback that time could not be expected so much even if the collation processing device was added.

Even in the remote batch method that processes a large amount of inquiry data all at once, the total running time of each collation processing unit becomes unbalanced, so the collation response time between each collation processing unit also becomes unbalanced and the collation response time becomes unbalanced. There was a drawback that it decreased.

The reason that the total matching running time of each matching processor is unbalanced is that the reference area of the query data depends on the characteristics of the query data, and when the characteristics of the input query data change, the reference area of the query data changes. This is because the matching running time of each matching processing device depends on the balance of the reference area of the inquiry data. Therefore, if the balance of the reference area is lost, the balance of the matching running time is also lost.

The present invention solves the above-mentioned drawbacks, and an object of the present invention is to provide an automatic collation system capable of shortening the collation response time of inquiry data in both the online real-time system and the remote batch system.

[Means for solving problems]

The present invention relates to a central processing unit including inquiry data distribution means for distributing input inquiry data, n (n is a positive integer) inquiry data files respectively storing outputs of the inquiry data distribution means, and n pieces. In the automatic collation method, the central processing unit comprises the master data file of n and the n collation processing devices for collating the content of the n reference data files with the corresponding content of the corresponding master file. By assigning the divided areas divided based on the hierarchical structure of n to the master files of n, each of the master files can be stored in the n master files so that the amount of data stored therein is substantially uniform. The reference data distribution means includes means for distributing data to the master file, and the reference data distribution means includes reference area information of the master file corresponding to the input reference data. The added characterized in that it comprises a means for providing this added queried data to the n pieces of reference data file.

[Action]

The central processing unit distributes data to each master file based on the characteristics of the inquiry data so that the amount of data stored in the n master files becomes substantially uniform. The inquiry data distribution means adds the reference area information of the master file corresponding to the input inquiry data, and adds the added inquiry data to n.
To each query data file. By the above operation, the reference data amount stored in each master file becomes almost uniform, so if the collation responses are performed in parallel, the processing can be performed in an even time and the time can be shortened. .

〔Example〕

Embodiments of the present invention will be described with reference to the drawings. First
FIG. 1 is a block diagram of an automatic collating device according to an embodiment of the present invention. In FIG. 1, the automatic collating device is a collation control as an inquiry data distributing means for distributing an input transaction file 21 composed of a magnetic disk or the like in which transaction data inputted from the outside and inquiry data inputted from the input transaction file 21 are distributed. The central processing unit 10 including the unit 11 and the inquiry data distribution unit 12, n inquiry data files 41 _{1 to} 41 _n for storing the outputs of the inquiry data distribution unit 12, and n master files 51 _{1 to} 51, respectively.
_n and n pieces of collation processing devices 31 _{1 to} 31 _n for collating the contents of the _n inquiry data files 41 _{1 to} 41 _n with the corresponding contents of the corresponding master files 51 _{1 to} 51 _n , and the candidate data are A candidate data file 61 to be stored is provided.

Further, the central processing unit 10 starts the collation of the collation processing devices 31 _{1 to} 31 _n according to the instruction of the collation control unit 11 to control the parallel processing of the collation, and also outputs the candidate data output from the collation processing devices 31 _{1 to} 31 _n. Collation processing device control unit 13 for input and collation processing device control unit
It includes a candidate data buffer 14 for storing the output of 13, and a candidate data output unit 15 for storing the contents of the candidate data buffer 14 in a read candidate data file 61 according to an instruction from the matching control unit 11.

Here, a feature of the present invention is that the central processing unit 10
Includes means for distributing data to each of the master files 51 _{1 to} 51 _n based on the characteristics of the inquiry data so that the amount of data stored in the _n master files 51 _{1 to} 51 _n becomes substantially uniform. In addition to the collation control unit 11 and the inquiry data distribution unit 12, the inquiry data distribution unit includes a divided area formed of a magnetic disk or the like for managing the divided areas of the master files 51 _{1 to} 51 _n which are storage locations of master data. Management file 22
And a temporary inquiry data file 23 in which inquiry data to which the contrast area information of the corresponding master file 51 is added are temporarily stored. Further, the inquiry data distribution unit 12 receives the input transaction file 21 according to an instruction from the collation control unit 11. Based on the contents of the divided area management file 22, the reference area information of the corresponding master file 51 is added to the inquiry data input from, and the added inquiry data is temporarily stored in the temporary inquiry data file 23 and the inquiry data file 41. It is to include means to give _{1 to} 41 _n .

The operation of the automatic matching device having such a configuration will be described. FIG. 2 is a diagram showing management of divided areas of a master file of the automatic matching apparatus of the present invention. FIG. 3 is a diagram showing the operation of the automatic matching apparatus of the present invention.

In FIG. 2, the large capacity master file 80 is an aggregate of master data. Reference numeral 81 denotes a large-capacity master file 80, which is a hierarchical structured file having a depth m according to the characteristic information of the master data. In the logical file 70, the large-capacity master file 80 is expanded into _α divided areas f _{1 to} f _α by hierarchical structure. The divided areas f _{1 to} f _α are respectively processed by the collation processing devices 31 _{1 to} 3 _1.
Evenly divided by the number of n collation processing devices up to 1 _n , and the divided regions f _{11 to} f _1n , f _{21 to} f _2n , ..., F _{α1 to} f
_It is composed of _αn .

The physical file 50 has N master files of master files 51 _{1 to} 51 _n , and is divided into regions f _{11 to} f of the logical file 70.
_{1n, f 21 ~f 2n, ...} , f α1 ~f αn are assigned to the n master file 51 of the sequential master file 51 ₁ to 51 _n.

Therefore, in the master files 51 _{1 to} 51 _n , the divided area f ₁
The divided areas obtained by dividing ~ f _α by the number N of matching processing devices are evenly allocated.

In Fig. 3, the inquiry data is a temporary inquiry data file.
The flow from 23 to the collation processing devices 31 _{1 to} 31 _n , how parallel processing of inquiry data is performed, and the flow of master data will be described. The dotted arrow 100 indicates the flow of inquiry data, the solid arrow 200 indicates the flow of master data, and the shaded area indicates the reference area of the inquiry data.

The operation will be described assuming that one inquiry data S is stored in the temporary inquiry data file 23.

Flow of Inquiry Data The inquiry data S stored in the temporary inquiry data file 23 is read by the inquiry data distribution unit 12 and distributed to the inquiry data files 41 _{1 to} 41 _n as the inquiry data S _{1 to} S _n , respectively. It is read by the collation processing devices 31 _{1 to} 31 _n . The inquiry data S _{1 to} S _n are respectively master files 51 _{1 to} 5
The 1 _n diagonally divided area is a control area, and the control area information is added.

Flow of master data Each collation processing device 31 _{1 to} 31 _n sequentially reads the master data stored in the shaded area in the divided areas of the master files 51 _{1 to} 51 _n based on the reference area information of the read inquiry data. .

Parallel processing operation of the same inquiry data Each collation processing device 31 _{1 to} 31 _n reads out the inquiry data from the inquiry data files 41 _{1 to} 41 _{n according} to an instruction from the collation processing device control unit 13 shown in FIG. _{1 to} 31 _n collate the same inquiry data at the same time.

〔The invention's effect〕

As described above, according to the present invention, it is possible to distribute the same inquiry data to all the collation processing devices and make the master collation amount of each collation processing device uniform, and to collate the inquiry data in proportion to the addition of the collation processing devices. There is an excellent effect that the response time can be shortened.

In addition, even if a large amount of inquiry data with any characteristics is input, the matching time of matching of each matching processing device will be uniform because the amount of master matching between the matching processing devices of the same inquiry data is uniform. There is an advantage that the matching response time can be shortened.

[Brief description of drawings]

FIG. 1 is a block diagram of an automatic collating device according to an embodiment of the present invention. FIG. 2 is a divided area management diagram of the master file of the automatic matching apparatus of the present invention. FIG. 3 is a diagram for explaining the operation of the automatic matching device of the present invention. 10 ... Central processing unit, 11 ... Collation control unit, 12 ... Inquiry data distribution unit, 13 ... Collation processing unit control unit, 14 ... Candidate data buffer, 15 ... Candidate data output unit, 21 ... Input Transaction file, 22 ... Divided area management file, 23 ... Temporary inquiry data file, 31 _{1 to} 31 _n …… Collation processing device, 41 _{1 to} 41 _n …… Collation data file, 50 …… Physical file, 51 ₁ 〜 51 _n …… Master file, 61 …… Candidate data file, 70 …… Logical file, 80 …… Large capacity master file, 81 …… Hierarchical structured file, 100 …… Dotted arrow, 200 …… Solid arrow, f ₁ ~ F _α , f ₁₁ ~ f _1n , f ₂₁ ~
_{f 2n, ..., f α1 ~f} αn ...... divided region, S, S ₁ ~S _n ...
... inquiry data.

Claims (1)

[Claims] 1. A central processing unit including inquiry data distributing means for distributing inputted inquiry data, n (n is a positive integer) inquiry data files respectively storing outputs of the inquiry data distributing means, and n. Master files and the above n
In an automatic collation system comprising n collation processing units for collating the contents of each inquiry data file with the corresponding contents of the corresponding master file, the central processing unit divides based on the hierarchical structure of the data. By allocating the divided areas to n master files, the data is distributed to each master file based on the characteristics of the inquiry data so that the data stored in the n master files are substantially uniform. The inquiry data distribution means includes means for adding reference area information of the master file corresponding to the input inquiry data, and providing the added inquiry data to the n inquiry data files. Characteristic automatic matching method.